obsidian-deploy-integration
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes project-specific metadata files (manifest.json, community-plugins.json, versions.json) and uses their content to drive repository actions via the git and gh CLIs.\n
- Ingestion points: Metadata extraction from manifest.json and modification of community-plugins.json in SKILL.md and references/implementation-guide.md.\n
- Boundary markers: None present; the skill treats local JSON content as trusted metadata for deployment steps.\n
- Capability inventory: git operations (push, tag, checkout, commit), gh operations (fork, pr create, release create), and npm operations (ci, build, version) in SKILL.md.\n
- Sanitization: None detected; the skill interpolates extracted metadata directly into shell commands and PR descriptions.\n- [COMMAND_EXECUTION]: The skill executes multiple shell commands using the git, gh, and npm CLIs in SKILL.md to manage project state, create releases, and submit pull requests. These operations are core to the skill's purpose as a deployment tool.\n- [DYNAMIC_EXECUTION]: The skill uses node -e and node -p in SKILL.md to execute inline JavaScript for JSON manipulation and version extraction. This is used for simple tasks using static logic provided in the instructions.\n- [EXTERNAL_DOWNLOADS]: The skill uses npm ci in SKILL.md and references/implementation-guide.md to install dependencies from the official npm registry during the plugin build process.
Audit Metadata